Determining the Effect of Some Biasing Parameter Selection Methods for the Two Stage Ridge Regression Estimator

Abstract

The use of biased estimation techniques is inevitable in connection with multicollinearity. Two stage ridge estimator is a pioneer biased estimator which is use to recover the problems that are originated from the multicollinearity. The noteworthy issue regarding two stage ridge estimator is selection of its biasing parameter. This article investigates several methods on selection of the biasing parameter of the two stage ridge estimator based on the works in the literature related to ridge estimator in a linear regression model. To demonstrate the best estimators of the biasing parameter, a Monte Carlo experiment is conducted. The utility of the proposed estimators of the biasing parameter for two stage ridge estimator is observed in terms of mean square error criterion.